Monobromoiron(III) ion. Equilibrium studies on the ... - ACS Publications

Ion Pair and Kinetic Studies on the Substitution and Reduction Reactions1. By DAVID W. CARLYLE and JAMES H. ESPENSON10. Received August 27,1968...
0 downloads 0 Views 773KB Size
MONOBROMOIRON(III) ION 575

VoL 8, No. 3, March 1969 INSTITUTE FOR

CONTRIBUTION FROM THE DEPARTMENT OF CHEMISTRY AND THE ATOMICRESEARCH, IOWA STATE UNIVERSITY, AMES,IOWA50010

Monobromoiron(II1) Ion. Equilibrium Studies on the Complex and on the Ion Pair and Kinetic Studies on the Substitution and Reduction Reactions1 BY DAVID W. CARLYLE AND JAMES H. ESPENSONIC

Received August 27,1968 Studies on iron(II1)-bromide complexes reveal two species-an inner complex FeBrZ+ and an ion pair Fe3+,Br--whose stabilities have been separately evaluated. At 1.6', /.t = 1.00 M , Q, = 0.034 f 0.003 M-1, and Qo = 0.20 M-1. The substitution rate constants for FeBr2+a t 1.6' and [H+] = 1.00 M a r e : aquation, 10.7 sec-1; formation, 0.34 f 0.03 M-1 Set-1. The reduction reactions of FeBr2+ and Fe3+,Br- by C r 2 +and E u 2 + have been studied. The values of kFeBr are 2 2 x 107 M-l sec-' for Cr2+and 1.3 X lo6 M-l sec-l for E u 2 +a t 1.6" and /.t = 1.00 M . The third-order rate constants under the same conditions are 3.9 X IO3 M F 2sec-l for C r 2 +and 5.8 X lo3 M - z sec-l for Eu2+. The mechanisms of the various reactions are discussed. The third-order reactions are best understood as reactions of the ion pair and the reducing agents.

Introduction The rate of reduction of the monobromoiron(II1) complex (H20)jFeBr2+by Cr2+ and E u 2 + is of considerable interest as one of a series of similar react i o n ~ . ~ The - ~ net equations for these two processes are

+ +

+ (Hz0)5CrBr2+ (I) + + Br- (11)

(Hz0)5FeBr2+ C r ( H ~ o ) e ~=+Fe(Hz0)e2+ (Hz0)bFeBr2+

=

F e ( H z o ) ~ ~ +Eu,,~+

One question of obvious interest in the latter reaction is whether electron transfer occurs by an inner-sphere mechanism involving a bridging bromide ion. We note that "direct" evidence of bridging is not likely, however, since either of the potential products resulting from Br- bridging, (H20)SFe1'Br+or Eu1I1BrZ+,would dissociate essentially upon its formation following electron transfer. The study of reactions I and I1 was complicated by the existence of two monobromo cations, the inner complex cited above and an ion pair, or outer-sphere complex, F e ( H ~ 0 ) 6 ~ + , B r - .T h e formation and aquation of the inner complex, shown in reaction 111, competed Fe(HeO)e3+

+ Br-

=

(HzO);FeBr2+f HzO

(111)

with reaction 11. Rate measurements under the conditions of interest were carried out for reaction 111, which had been studied previously by Matthies and Wendt5 The stability quotients for the inner complex, reaction 111, and the ion pair, reaction IV, are designated Fe(H20)e3+4- Br- = Fe(H20)e3+,Br-

(IV)

Qi and Qo, respectively. The spectrophotometric study of Lister and Rivington6 on equilibrium solutions yielded a single equilibrium quotient Q, the sum of Qi and Q0.7 (1) (a) Work was performed in the Ames Laboratory under the adspices of the U. S . Atomic Energy Commission; Contribution No. 2398. (b) Based on the Ph.D. thesis of D. W. C., Iowa S t a t e University, M a y 1968. (c) Fellow of t h e Alfred P. Sloan Foundation. (2) D. W. Carlyle and J. H. Espenson, J. A m . Chem. Soc., 90, 2272 (1968). (3) G. Dulzand N. Sutin, ibid., 86, 829 (1964). (4) D. W. Carlyle and J. H. Espenson, ibid., 91. 699 (1969). ( 5 ) P. Matthies and H. Wendt, Z. P h y s i k . Chem. (Frankfurt), SO, 137 (1961). ( 6 ) M . W. Lister and D. W. Rivington, Can. J . Chem., 38, 1603 (1955).

In this study we have evaluated the following quantities pertaining to the iron(II1) complexes: (a) the separate equilibrium quotient Qi for (H20)jFeBr2+ by two independent methods, (b) the value of Qo by difference, after a value comparable to that of Lister and Rivington6 was obtained for the sum, (c) some uvvisible spectral data for both complexes, from measurements on solutions in which ion-pairing reaction IV was a t equilibrium, but where significant inner complex had not yet formed, (d) the aquation and formation rates of the inner complexation reaction, eq 111, (e) kinetic studies of the rates of reduction of the inner complex by Cr(I1) and Eu(I1) as in reactions I and 11, and (f) kinetic studies on the rates of reduction of the Fe3+,Br- ion pair by the same reducing agents, as shown in the reactions l ? e ( H ~ o ) ~ ~ + , -k B rCr(Hz0)s2+ = Fe(H20)eZt Fe(H20)e3+,Br-

+ (H20)jCrBrZt + HzO

(V)

+ EU,,~+= Fe(H20)e2++ Br- + EuaQ3+ (VI) Experimental Section

Materials.-The preparation, storage, transfer, and analysis of chromium(II), europium(II), lithium, and iron(II1) perchlorate solutions have been described.2z4 The source of bromide ion was reagent grade hydrobromic acid. Traces of bromine were removed by treating HBr stock solutions with a small concentration of iron(I1) perchlorate, the latter prepared by reducing a solution of the iron(II1) salt with amalgamated zinc. The HBr stock solutions to be mixed with Cr2+or E u 2 +solutions were not pretreated with Fez+, but the concentration of each diluted Cr2+ or Eu2+ solution was always analyzed after mixing with HBr. Reagent grade perchloric acid was used without purification. The water used throughout was prepared by twice redistilling laboratory distilled water from alkaline permanganate solution in a tin-lined Barnstead still. Measurements.-A Cary Model 14 recording spectrophotometer with a thermostated cell holder* was used for spectral measurements. The rates of rapid reactions were measured using the stopped-flow apparatus described previously.g In experiments where product analyses were to be made, the eight-jet mixing (7) Only one value is obtained, and i t is the unweighted sum indicated, irrespective of the relative abundances of the two complexes or of t h e contribution each species makes to the property measured: E. L. King, J. H. Espenson, a n d R . E. Visco, J . Phys. Chem., 63,755 (1969). (8) J. H. Espenson, J. Am. Chem. Soc., 86, 5101 (1964). (9) D. W. Carlyle and J. H. Espenson, Inovg. Chem., 6, 1370 (1967).

AXD JAMES H. ESPENSOX 576 DAVIDw. CARLYLE

chamber and syringe assembly of the flow instrument was used. This procedure was found to be essential for obtaining reproducible product distributions from very rapid reactions. All of the rate and equilibrium studies described here were carried out a t ionic strength 1.00 M , maintained with LiC1O4. The substitution rates of the inner complex were :tudied using the stopped-flow technique in experiments a t 4050 A, by mixing Fe3+ with Br-. These experiments measured the approach to equilibrium in reaction I11 and in principle would allow eualuation of the forward and reverse rate constants. In fact, the reverse (aquation) rate so dominated the experiment, as i t did also in the earlier work of Matthies and Wendt6 on the same reaction under different conditions, that the forward (formation) rate was not evaluated by such measurements. An indirect approach to measure the formation rate was employed, by making that step a major rate-determining component upon mixing a solution of Fe3+ (in HC104) with one containing CrZ+ and Br-. The FeBrZ+so formed was quantitatively converted to CrBr2+, which was separated by ion exchange using Dowex 50\V-X8 resin according to the procedure d e ~ c r i b e d . ~ The equilibrium quotient for the inner complex, Qi, was obtained by two methods-the quotient of forward and reverse rate constants for reaction 111, and an analysis for the inner complex FeBrZf in equilibrium solutions based upon its quantitative conversion t o CrBrZ' upon reaction with Crzi. The CrBr2+ so obtained was separated and analyzed as before. The value of the gross formation quotient, Q, was obtained spectrophotometrically under the conditions of interest in a procedure similar to that of Lister and Rivington.6 The difference of Q and Qi gave Qo, the stability quotient for the proposed ion pair. Experiments in which Fe3+ and Br- were mixed using the stopped-flow apparatus revealed considerable uv absorption formed within a time where an inappreciable concentration of the inner complex could have formed (the latter conclusion based upon the rates of reaction 111). The instantaneous absorption is attributed t o the ion-pairing process, reaction IV. The rate constants for the Fe3+-Br- reduction reactions were measured by the stopped-flow procedure by mixing Fe3' solutions with Cr2+ or EU2+,bromide ion being exclusively in the latter solution. For Br-, unlike other ions,4 the rates of (111) are sufficiently competitive that corrections for formation of FeBr2+ were necessary, but such interference by the inner complex was small. The reactions of FeBr2+ were studied with iron(II1) and bromide ions in solution together hefore mixing with the reducing agents. Competition of the other reactions was quite severe during the study of FeBr2+ Eu2+, and three different procedures were used to circumvent these difficulties, as described in a later section.

+

Results Aquation Rate of (H,0)sFeBr2+.-An absorption increase a t 4050 8,A, for the inner complex, occurred at a rate measurable by the stopped-flow procedure, upon mixing Fe3+ and Br-. The rate of formation of FeBr2+ according to reaction I11 is governed by eq 1, d[FeBr2+]/dt = k'f[FeaT][Br-]

- k',q[FeBr2+]

(1)

where k'f and k',,, which are functions of [H+],are related to Q1: Q1 = k'f/klaq. According to this expression, the approach to equilibrium in reaction I11 should follow pseudo-first-order kinetics under conditions where [Br-1 remained substantially constant. The expression for the apparent rate constant, derived by the usual procedure^,^ lo is that given by eq 2. The - d l n ([FeBr2+], - [FeBrZC]}/dt=

k,,,

=

k'f[Br-I

+ k'aq

(2)

Inorganic Chemistry appropriate rate plotsg-" followed the expected firstorder behavior as evidenced by linear rate plots to a t least 90% completion. Repetitive measurements on the same set of solutions generally agreed to within 2%. Also in accord with eq 2 , the values of k,,, were not dependent on the iron(II1) concentration. The values of kapp, covering the range 0.014 5 [Br-] 5 0.50 M, are summarized in Table I. Over the entire range of bromide concentrations k,,, remained constant within experimental error, indicating k'f [Br-] was very small relative to k'zq. These results give k'aq = 10.7 f 0.5 sec-l (l.So, 1.00 M H + , p = 1.00 M), and in a single experiment a t 15.8", k'aq = 55.6 sec-I. TABLE I RATE CONSTANTS FOR APPROACH TO EQUILIBRIUM IN REACTION 111"

a

Temp,

lP[Fe:III) IO,

O C

F

103[Br-]o,

&pp,

M

sec-1

1.6 14.2 - 14.2 25.6 1.6 25.6 1.6 3.00 133 1.6 2.00 200 400 1.6 1.oo 1.6 0.800 500 15.8 51.2 80.0 Conditions: [H+] = 1.00 M, p = 1.00 M,X 4050

10.4 11.3 10.9 10.7 11.1 10.5 55.6

A.

The time needed to reach a stable absorption spectrum in Fe3+ solutions containing bromide ions was ca. 0.5 sec; subsequent to this no spectral changes were noted over the entire uv-visible regions even on standing several days. The rate constants are t y p i ~ a l ~ , ~ ~ ' ~ of substitution in the inner coordination sphere of Fe(H20)63+and can be assigned to the inner complex, which is an assignment consistent with the kinetic derivations outlined above. Formation Rate of (HzO)6FeBr2+.-The very low stability of FeBr2+prevented a direct evaluation of the formation rate constant k'f from the rate of approach to equilibrium. The following indirect procedure was devised. The complex FeBrZ+ oxidizes Cr2+far more rapidly than does either of the species Fe3+ or Fe3+,Br-. An appropriate choice of concentration conditions permits the following sequence of reactions to be studied k'f

Fe3+ f Br-

--t

Cr2+

FeBr2++CrBr2+ fast

The rate of Fe3+loss by this sequence accounts for an appreciable portion of the total when a solution of Fe3+ (containing no bromide) is mixed with a solution containing Cr2+ and Br-. Competing with this pathway are three additional reactions consuming iron(II1) : Cr2+ Fe3+, Cr2+ FeOH"', and Cr2+ Fe3+ f Br-. The total rate of disappearance of iron(II1) is given by

+

+

+

(10) D.R.Seewald and N . Sutin,lnovg. Chenz., 2,643 (1963). (11) E.A. Guggenheim, Phil. M u g . , [712, 137 (1926). (12) F. P.Cavasino, J . P h y s . Chem., 72, 1378 (1968).

MONOBROMOIRON(III) ION 577

VoZ. 8,No. 3, March 1969 -d[Fe3+]/dt = k'i[Fe3+1[Br-I f [Fe3+][CrZ+]{ki

+ ( R z / [ H + ] )+ km[Br-] )

(3)

in which kl and kz, representing the respective reductions of Fe3+ and FeOH2+ by Cr2+, are known from previous s t u d i e ~ , 3and ? ~ k B r is derived from experiments to be presented subsequently in this paper. At 1.6" and p = 1.00 M (with LiClOd) the values are4 k1 = 250 M-I sec-l, kz = 800 sec-l, and k B r = 3.9 X l o 3 sec-l. The first term in eq 3 was made as appreciable as possible in these experiments by maintaining a low Cr2+ Concentration. The iron(II1) concentration was lower still, however, so that [Cr2+ ] remained essentially constant. Under such conditions the disappearance of iron(II1) is expected to follow pseudo-first-order kinetics. According to eq 3 the rate constant so evaluated (-d In [Fe3+]/df) should be the quantity given by k

=

k'f[Br-I -I-[Cr2+1,,{k~f (ke/[H+l)

+ k~rIBr-1)

(4)

Experiments under four sets of concentration conditions were carried out and each gave the linear first-order plot predicted. Table I1 summarizes the results of the

RATECONSTANT k'f

[Br-1, A4 0.150 0.300 0.350 0.500 a

TABLE I1 Cr2+ Br-

FRohi

104[[Crzt]a, Ad

104[Fe"'lo,

1.66 0,925 1.80

0,200 0.150 0.200 0.200

1.15

F

+

+ Fe3

+

STUDIES"

kobsd, sec-1

k'i, M-1 sec-1

0,307 0.30 0.515

0.35 0.37 0.30 0.364 0.34f0.03

0.497

Av Conditions: l.6", [H+] = 1.00 M , p = 1.00 M .

experiments. A rearrangement of eq 4 yields an expression for k'f; with k measured and the other quantities known, a value of k'f can be calculated. The measured rate constant k is higher than the value computed were k'r = 0 by 17, 37, 20, and 37%, respectively, in the four experiments. Thus the first term in eq 3 can be made quite significant, but it was not so large as to predominate in the entire experiment. Table I1 summarizes the values found for k'f, the average being 0.34 0.03 M-I sec-I (1.6", [H+] = 1.00 M , y = 1.00

*

MI. Stability Constant of (H20)5FeBr2 +.-One value of is calculable from the measured k'f and ktaq values: Qi = k'f/krag. The results obtained above give Qi = 0.032 f 0.003 M-l a t 1.6" and p = 1.00 M. The value of Qi was also determined under the same conditions by a completely independent technique. Solutions of FeBr2+,in equilibrium with Fe3+and Br-, were mixed (using the stopped-flow mixing chamber) with excess Cr2+. The resulting CrBr2+was quantitatively separated by ion-exchange c h r ~ m a t o g r a p h y . ~ These experiments always had [Fe(III)]o > [Cr2+Io> [FeBr2+Io,so that no Cr2+remained unoxidized. (This is an important point, since the reaction Crz+ 0 2 in the presence of Br- produces CrBr2+, in amounts sufficient t o vitiate such results.) Qi

+

The total [CrBr2+]in the product solution was assumed to form, quantitatively, in the following ways: (1) very rapid oxidation of Cr(I1) by the FeBr2+present a t the instant of mixing, (2) oxidation of excess Cr(I1) by the third-order kg, path (described later in this paper), and (3) oxidation of excess Cr(I1) by FeBr2+ formed after the initial FeBr2+ was reduced. The concentration of FeBr2+ present a t the instant of mixing was computed by an iterative procedure, as follows. A value for Qi was assumed permitting calculation of the concentrations of all of the species in the mixed solution immediately after reduction of the FeBr2+ initially present. These concentrations, together with the previously measured rate constants-kl, kz, kBrand the value for k'f consistent with k'aq = 10.7 sec-1 and the assumed value for Qi, were used to calculate the concentration of each species as a function of time, until the Cr2+ was consumed. A Runge-Kutta numerical method similar to that described by Espenson and Parker13 was used to solve the differential equations involved in this computation. The yield of CrBr2+ so calculated was compared to the observed, a new Qi was selected, and the computation was repeated until the two concentrations agreed. Three experiments were performed. The data together with the Qi values so computed are summarized in Table 111. On the basis of the sensitivity of the convergence, we estimate the precision of Qi by this method a t roughly 10-20%. The average of these three experiments gives Qi = 0.039 f 0.006 M-l, which can be compared with Qi = 0.032 f 0.003 M-' computed from the rate constants. Spectrophotometric Measurements on Fe3+ BrSolutions.-Absorbance measurements on solutions of iron(II1) containing various high bromide concentrations, with 1.00 M H+, were made a t 4050 A, the reported maximum wavelength for the bromo complex. The absorbance values (0) for an optical path of b cm were converted to an apparent molar absorptivity of iron(III), i, by the relation i = (D - Do)/b[Fe(III)], where Do represents the absorbance with HC104 in place of HBr. The values of 5 observed as a function of bromide ion concentration and temperature are summarized in Table IV. Were a single iron(II1) bromide complex formed, eq 5 would apply,14 where €1 represents the molar absorp-

+

1 =

€1

- (l/Q)(;/[Br-])

(5)

tivity of the presumed single complex and Q represents its stability constant. Linear plots were obtained when the data of Table IV were fit to eq 5. The very small degree of association did not permit a very severe test, nor were the parameters very precise. The values a t each temperature are: 10-2~1(M-I cm-I) = 7.0, 8.7, and 7.9 a t 1.6, 15.8, and 25.0°, respectively; Q = 0.23, 0.26, and 0.37 M-' a t the same three temperatures. The direct evaluations of Qi described above gave Qi = 0.032 f 0.003 and 0.039 f 0.006 J 4 - I a t 1.6" com(13) J. H. Espenson and 0. J. Parker, J. A m . Chem. Soc., 90, 3689 (1968). (14) T. W. Newton and G. M. Arcand, i b i d . , 75, 2449 (1953).

578 DAVID'IV. CARLYLE AND

H. ESPENSON

JAMES

EVALUATIOK OF Qi F

0.0300 0.0250 0.0300 Conditions: 1.6', 1.00 d l Qi giwn for each experiment.

p = 1.00

M.

1O4[CrBrZt], M (obsd)

104[FeBr2 + ] ~ , b ' ~

Qi,

M -1 M . (calcd) 5.75 3.60 0.032 2.13 3.85 3.59 0.047 3.14 3.89 3.18 0.037 2.47 Concentrations after mixing but before reaction. c Computed using the value of M

0.111 0.133 0.111

H-,

TABLE 111 CrBr2+YIELDEXPERIMENTS~

BY

104 [CrZ+lo:

[Br-I,* AI

[Fe(III)

Inorganic Chemistry

-d[Fe3+]/dt = kB,[Fe3+][M2+][Br-] (7) TABLE 11' APPARENTMOLARABSORPTIVITYOF IRON(III) AS A FUSCTION purpose bromide ion was added to only the reducing OF BROMIDE COXCENTRATION AND TEMPERATURE^ 10~[Fe(III)I, F

[Br-I,

>vi

50.0 0.0200 7.28 0.0799 7.28 0.133 7.28 0.173 7.28 0.200 3.64 0.533 0.800 3.64 a Conditions: [H+] = p

,-----; 1.6'

, -M -1 cm-1--------. 15.8'

7.52 10,88 13.2 18.2 21.1 29.1 26.9 37.4 30.4 42.6 77.7 107 150 109 = 1.00 M , X 4050 A.

25.0°

14.0 23.6 37.6 48.2 54.6 131 181

pared with Q = 0.23 M-' from the spectrophotometric data. This discrepancy suggested that a second species, a labile ion pair, was important under the conditions of this study, as described in reaction IV. I n this event the appropriate relation for is

Independent evidence was obtained for a species formed immediately upon mixing Fe3+ and Br- by stopped-flow measurements a t 1.6".I n the wavelength range 2500-3300 A, the absorbance of the solution freshly mixed in the flow apparatus was notably higher than that computed for the separate components Fe3+ and Br-. The immediate absorbance was, however, lower than the equilibrium value reached after occurrence of inner complexation reaction 111. The data were not very precise, but they did serve to establish the presence of a species other than FeBr2+, which had a strong uv absorption. No such instantaneous absorbance enhancement was noted at 4050 A, making it possible to reconsider the d a t a of Table IV. I n the limit where E,